Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) has become an important tool in the diagnostic evaluation of gastrointestinal tract lesions and other organ sites such as mediastinal and intra-abdominal lymphadenopathy, pancreatic masses, liver masses, left adrenal masses and gastrointestinal submucosal lesions. It provides crucial information that can have tremendous impact on patient management. FNA is typically performed using a 22- or 25-gauge needle with a stylet under EUS guidance. The lesion is punctured with a stylet in place in the needle. After withdrawal of the stylet, the needle is moved to and fro within the lesion and this process is repeated for each needle pass. It is currently believed that the use of a stylet for EUS-FNA improves the quality of specimens by preventing the tip of the needle being clogged up with tissue and hence enhances the diagnostic yield of specimens obtained. However, there are no data demonstrating clearly that the use of a stylet improves the yield of EUS-FNA. The reason why this question is important is because the use of a stylet during EUS-FNA is cumbersome, time and energy consuming and increases the costs of EUS-FNA needle systems.

In this prospective randomized controlled trial, patients referred for EUS-FNA of mediastinal and intra-abdominal lymphadenopathy, pancreatic mass, liver mass, left adrenal mass and gastrointestinal submucosal tumors will be included. FNA will be performed with a 22-gauge needle under EUS guidance using suction with a 10 mL syringe by two experienced endosonographers. The technique to be used for fine needle sampling i.e. with a stylet in place or without a stylet for each FNA pass will be assigned by using a preprinted randomization scheme obtained from a sealed envelope and clearly documented. Each lesion will be sampled for a minimum of four needle passes. The pathologists providing the final interpretation will be blinded to technique of EUS-FNA (with or without stylet). The degree of cellularity, contamination, amount of blood, adequacy of sample, frequency with which a positive diagnosis is made will be compared between the two groups (EUS-FNA with stylet vs. EUS-FNA without stylet). The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of each technique when compared to the final diagnosis will be calculated. Inter-observer agreement among cytopathologists will be assessed for specimens obtained from EUS-FNA with stylet and for those obtained from EUS-FNA without a stylet.

Various techniques have been described to optimize accuracy, efficiency, and quality of EUS-FNA specimens. FNA is typically performed using a 22- or 25-gauge needle with a stylet under EUS guidance. The lesion is punctured with a stylet in place or slightly withdrawing the needle. After puncture, the stylet is pushed out of the needle tip and then the needle is moved to and fro within the lesion and this process is repeated for each needle pass. It is currently believed that the use of a stylet for EUS-FNA helps prevent clogging of the needle by gut wall tissue, which could limit the ability to aspirate cells from the target lesion. This may improve the quality of specimens and hence enhance the diagnostic yield of specimens obtained. This is a logical assumption, but there are no data demonstrating clearly that the use of a stylet increases the yield of EUS-FNA. At the present time, it is recommended that the stylet is re-inserted back into the needle prior to each FNA pass. The use of a stylet during EUS-FNA is cumbersome, time and energy consuming and increases the costs of EUS-FNA needle systems. In some circumstances, the stylet may actually make EUS-FNA very difficult as it may be impossible to advance or remove the stylet once the target has been punctured. This tends to occur when the echoendoscope or the needle is bent and a large (19 gauge) needle is being used. In addition, the data comparing the effectiveness of EUS-FNA with stylet to FNA without stylet is limited. Paquin et al compared the adequacy, the bloodiness, and the yield of FNA samples obtained with a stylet to FNA without a stylet. In this study, the use of stylet for EUS-FNA was associated with a reduced specimen adequacy and more bloody passes. 13 Thus the use of a stylet for EUS-FNA is questionable and needs further investigation. If the diagnostic yield, adequacy and quality of specimens obtained by EUS-FNA without a stylet is found to be equivalent to that with a stylet, this could potentially make a strong case for not using a stylet and thus making the procedure easier, more time- and cost-efficient. The hypothesis and specific aims of this prospective randomized controlled trial are as follows:

First hypothesis: There is no difference in the degree of cellularity, contamination, and amount of blood in samples obtained by EUS-FNA with and without a stylet Specific Aim #1: To compare the degree of cellularity, contamination, and amount of blood in samples obtained by EUS-FNA with and without a stylet

Second hypothesis: There is no difference in the diagnostic yield of malignancy in specimens obtained by EUS-FNA with a stylet compared with EUS-FNA without a stylet.

Specific Aim #2: To compare the diagnostic yield of malignancy in specimens obtained by EUS-FNA with and without a stylet.

Third hypothesis: An acceptable level of inter-observer agreement exists among cytopathologists in the assessment of specimens obtained from EUS-FNA with stylet and EUS-FNA without a stylet.

Specific Aim #3: To assess the inter-observer agreement among cytopathologists in the evaluation of specimens obtained from EUS-FNA with stylet and specimens obtained from EUS-FNA without a stylet.

Inclusion Criteria:

• Age greater than 18 years
• Presence of mediastinal or intra-abdominal lymphadenopathy, solid pancreatic mass, left adrenal mass, gastrointestinal submucosal lesions or liver mass confirmed by at least a single investigational modality - CT scan, magnetic resonance imaging, endoscopy.
• Capable of providing informed consent

Exclusion Criteria:

• Severe coagulopathy (INR > 1.5) or thrombocytopenia (platelet count < 50,000)
• Lesion unable to be sampled due to the presence of intervening blood vessels
• Results of EUS-FNA would not impact patient management
• Inability to provide informed consent